1.3.15 (and earlier): cdrecord fifo bug

thomas tom@huno.net
Wed Nov 20 16:32:00 GMT 2002


hello,

using cygwin 1.3.15-2 (and earlier versions, i will elaborate later) to
compile cdrecord 1.11a39 (and earlier) makes it produce a buffer
underrun immediately after writing has begun. fifo will go down from
100% straight to 0% in about 1-2 seconds. jörg schilling thinks it might
be a scheduling bug in cygwin.

here is howto reproduce this error: grab any of the binaries from
http://www.geoshock.com/cdrtools/ (or compile your own of course) and
burn a cd, using this command f.e.:

mkisofs -J -R * | cdrecord -v -speed=24 -

you can add -driveropts=burnfree if your writer supports it, that will
save the cd-r, but fifo will stay at 0% during the whole time.

it will happen every time, i tested it on 2 computers with two different
writers. now comes the interesting part: i found some older cygwin
binaries from 1.1.18 (cygwin1.dll, /usr and some binaries), mixed them
with the 1.3.15-2, compiled cdrecord 1.11a39 with it and fifo works
flawlessy! not a single buffer underrun up to now. that means the bug
was introduced sometime between 1.1.18 and now (i never used cygwin
before).

thomas


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/



More information about the Cygwin mailing list